The Definition of Gender in Competitive Athletes – Edge of Growth and Maturation

The definition of gender in competitive athletes

A live conversation between

and

With comments by Paul Saenger, Jan Lebl, Jan-Maarten Wit, Ken Ong, Ze’ev Hochberg, Lyne Chiniara and Michael Ranke

Transcribed by Lyne Chiniara

Ivo Arnhold: As pediatric endocrinologists, we help individuals with differences (or disorders) of sex development (DSD). The DSDs deal with several challenges, among others their classification to participate in competitive sports. To discuss this issue, we are fortunate to have Dr. Martin Ritzén, Professor Emeritus at the Karolinska Institutet, Stockholm, Sweden, a well-known expert on the area of DSDs and an adviser to international sports authorities such as IAAF and IOC.

In almost all sports, athletes are divided into male and female categories. Subjects with DSDs must also compete accordingly. How can one preserve the right to compete and privacy of these individuals and maintain a fair level towards other competitors?

These are some main points we would like you to address: Why are men stronger and faster than women? Are individuals with DSDs overrepresented in some sports? Which DSDs and other causes of hyperandrogenism confer an advantage in sports and why? Doping vs. penalizing natural attributes. What are the present regulations for individuals with DSDs (and transsexuals) to participate as women? Different points of view and future trends.

Martin Ritzén: Basically all competitive sports have two different categories of participants: Female and Male. The only exception that I know about is for equestrian competitions. You can ride a horse as female or as male, in the same category. Five years ago, for the first time in history, one of the very famous horse races in Australia was won by a female, and that made headlines all around Australia. But otherwise the first thing to decide is whether the competitor is female or male.

The reason for this division is that in events that require speed and strength, males are generally superior to women. Without this categorization, women would stand no chance of winning World or Olympic championships. This is evident from comparing the results at world championships in running events; elite female competitors would end up far behind the males. The female gold medallist generally ends up at around 50th place if competing in the male category. The time difference is about 10-12%, in running time or other sports events. This is true for most sports event.

There has been a longstanding debate in how to define a person that is eligible to compete in the female category. In the 1930s, there were some publicized cases of women who won Olympic medals. For instance, Stella Walsh, of Polish origin, was named “the fastest woman in the world”. Her nickname was «Stella the fella», which indicates she was thought to be quite masculine. Unfortunately, she was killed in a traffic accident, and the autopsy revealed she had an intersex condition, but was raised and lived as a female.

Dora Ratjen was a German high jump competitor in the 1936 Olympics who was quite masculine. She ended up 4th in the Olympic Games in Berlin. Afterwards a rumour was spread that she was a masquerading male, who was competing to show the superiority of the Nazi regime. She later changed from female to male sex, and again this was used to say she was masquerading as a female. This was probably not true; rather, she probably had a DSD condition.

The separation of athletes into female and male categories requires a clear definition of the two. Since males are superior to women, most of the discussions aim at defining who is eligible to compete in the female category – how masculine can a woman be and still compete as a woman? The initial concerns were that some men might masquerade as females in elite competitions. This proved to be very rare. But suspicions that some successful female athletes were male persisted.

The risk of masquerading males as a female was brought up in 1948. The first athletic body to request documentation of somatic sex was the British Women´s Athletes Association. The sports authorities then decided to use “objective” proofs of femaleness. The first method was the “nude parade”; the women were requested to show themselves to a panel of gynaecologists. This was soon abandoned in favour of sex chromatin testing; a negative result, suggesting XY rather than XX karyotype, would disqualify the athlete. This was used for a number of years until this also proved inadequate. For example, women with androgen insensitivity were prohibited from competing as females, in spite of the fact that their androgen insensitivity made them even less masculine than healthy women. A famous case was the one of the Spanish athlete Maria Patino; she wrote a book about her situation as having CAIS. She underwent the sex chromatin test and first passed, because it is an unreliable test, but at a different competition, she had forgotten her certificate and went through the test again; this time she was disqualified because she did not have two X chromosomes. She was completely female because of CAIS, and you all know that having higher levels of testosterone does not make her more competitive because her body does not respond to the androgens. Her fiancé left her, she was stripped of all her previous medals. Later, when her diagnosis was revealed, she got them back.

Then, SRY testing had appeared; you could then prove that this person had a Y chromosome, and that worked for a while. However, in the Atlanta Olympic in 1996, a study showed that when all female athletes were tested for SRY, eight were SRY positive, but none of them was virilised. Some of them were gonadectomised years before, and still were very good at sports. One of them was a patient of mine; she had PAIS, with gonads removed in the neonatal period. Did she still benefit from having a Y chromosome 20 years later with no excess testosterone? This indicates that yet unknown factors other than the current testosterone levels may be important for excellence in sports.

When also this method failed to define eligibility to compete in the female category, IAAF and IOC were facing no rules at all when in 2009 Caster Semenya, then 17 years old, won the World Championship in 800 m running, far ahead of all others. She had a quite masculine appearance, big muscles, masculine gait, which prompted some of her competitors and news media to challenge her female sex. Her competitors complained that they did not want to compete with someone with a male body. In the newspapers, questions were raised whether she is a man or a woman. The spokesman of IAAF publicly said that they would “determine her true sex”, unaware of the fact that sex identity cannot be determined by any laboratory test. Only the person herself/himself can define her/his sex identity. That made me furious as she could have been one of my patients. I protested publicly in Sweden and was called for a meeting in Monaco with the IAAF. This organization realized that it was unacceptable to have a public discussion on anyone´s sex identity. In 2009 there were no rules because all previous tests had been abandoned. I suggested testing for testosterone, as it is a key factor.

IAAF set up a small group with the task to formulate new regulations on eligibility to compete in the female category, and I was asked to join this group. My first task was to list a number of DSD conditions that might confer an advantage over other women. That proved impossible; the individual variations within all the diagnostic groups are too large.

Instead, I searched for the physiological differences between male and female athletes that make males perform better than women.

What makes testosterone a doping agent in sports? Testosterone is well known to increase muscle strength and speed. Testosterone administration increases haemoglobin levels in blood, with marked effect on oxygen transport. In some studies, testosterone administration has an effect on CNS by increasing aggressiveness, which might improve performance in elite sports. Therefore, administration of anabolic/androgenic steroids is considered a severe offence to doping regulation and is punished by harsh penalties.

What about testosterone levels? Reference values for serum testosterone for healthy women reach up to 2.7 nmol/L (with immunoassays), while the lower reference limit for healthy men is about 9 nmol/L. In other words: a woman must reach 5 SD above the mean for healthy women in order to get into the male range. I suggested to use the lower range of the male normal range as a cutoff level; and after a long debate this was accepted (2011-2017). Hyperandrogenism of that magnitude can be reached in some DSDs, adrenal tumours and plain doping. Women with PCOS will rarely reach the male range of serum testosterone.

Are women with DSDs overrepresented in elite sports? Yes. Women with PCOS are overrepresented in successful women in sports. When Bermon et al measured testosterone levels in all 855 female athletes in the 2011 World Championships, 10 had previously unknown DSD with a mean serum testosterone concentration well into the normal male range. In the Atlanta Olympic games, when all participating women underwent SRY analysis, 8 women were found to be SRY positive, giving an incidence of 1/140, far higher than in the general population. It is remarkable that most of these women had their testes removed at the time of testing.

Which DSDs seem to confer the greatest advantage in female sports? 5-alpha reductase deficiency is the most common cause of hyperandrogenism among elite female athletes. Most of the cases that so far have come to the attention of IAAF come from African countries with limited access to medical, and especially neonatal, diagnostic facilities. 5α-reductase deficiency type II (5αRD) is the most common diagnosis, but a few have partial androgen insensitivity (PAIS). If women with this latter condition have retained some degree of androgen sensitivity, it has been argued that these women would benefit from their high testosterone levels. However, I am of a different opinion: A new-born child with PAIS that has so little masculinization of their external genitalia that they were assigned female sex will carry this poor androgen sensitivity with them into puberty and adulthood. Therefore, even if they prove to have high testosterone levels in blood, they should be allowed to compete as women.

Congenital adrenal hyperplasia (CAH) due to 21-hydroxyase or 11-hydroxylase deficiency might cause high levels of testosterone if untreated or undertreated. So far, no such athletes have been detected during anti-doping screening. Finally, women with true hermaphroditism, born with ambiguous genitalia might further virilise in and after puberty. Since they are assumed to respond normally to high testosterone levels, they would fall into the same category as 5αRD.

The present regulatory rules for women with extreme hyperandrogenism are that women with testosterone levels above 5 nmol/L should be subjected to investigation. If doping can be made probable, antidoping rules will be applied. If doping is excluded, the preliminary results of endocrine screening will be reviewed by at least two members of a standing expert panel. They will make recommendation to the IAAF medical commission to proceed or not with a full DSD work-up at one of the five internationally accredited tertiary centres for DSD diagnosis. The chairman of the expert panel advises the IAAF medical commission on whether the athlete should be required or not to suppress her testosterone levels by medical means, and the modes of follow-up during suppression. If the athlete accepts suppression, she will be monitored in order to prove that her testosterone remains below 5 nmol/L for at least 6 months before she is allowed to compete again in the female category. It is known from lay press that Caster Semenya has very high testosterone levels in blood.

Paul Saenger: What was her diagnosis?

Ritzén: Her medical diagnosis has not been publicly revealed.

Saenger: Was she CAIS?

Ritzén: It cannot be, because she would not have masculinized.

Audience: Then I wonder why testosterone should be a parameter to distinguish between male and female, because it would not catch the entire androgen insensitivity syndrome.

Ritzén: Let’s first go back to testosterone levels: if you have combination of high levels of testosterone and masculinization, it cannot be CAIS. If you have PAIS, raised as female, you then have very low androgen sensitivity. Provided that you were accepted as a girl at birth, maybe with some mild clitoromegaly, and then raised as a girl, you would probably maintain the same low degree of androgen sensitivity in puberty and later. So, I doubt personally, but I don’t have any scientific data backing this up, that anyone with PAIS raised female would ever have a benefit from testosterone. I base this on the few male patients with PAIS that I have treated outside the sports world. Even very large doses of testosterone had little effect on the appearance of their genitalia. I don’t know if anyone of you has had similar experience.

Anyway, during the seven years that I was chairman of the expert panel for the IAAF (2011 – 2018), 15 cases brought up for evaluation and diagnosis: They were all from Africa, most with 5αRD (all but 2). I think this is easy to explain; they were born in villages, with low levels of medical care. We do not know the level of virilisation at birth, but they passed as females. But at puberty, when the testes start to make testosterone they start virilising.

Saenger: If the testosterone regulation on classification still holds, this problem in the South African woman is still unresolved, and I find it disheartening that the athlete continues to appear in competitions. It is interesting that she only is doing the 800 m, apparently, this advantage gets washed out if she does shorter periods, this is unheard off. I understand why the 2-3rd athlete would complain.

Ritzén: In the official papers in 2011, she was asked to comply with the regulations that said that such high levels of testosterone, you have to reduce it for at least 6 months before she can compete again as a female. The simple way would be with the oral contraceptive pill, which would lower LH and FSH and thus lower testosterone levels. But between 2011 and 2015 she lost most international races. Then in 2015, the regulations that were introduced in 2011 were challenged by Caster Semenya’s lawyers in the court of arbitration in sport (CAS). CAS acts as a court, where an athlete can complain about the management of their respective sports authorities. In 2015 an Indian sprinter Dutee Chand complained to CAS over the way the Indian sports organization had prevented her from competing, due to her high testosterone levels in blood. Lawyers representing Ms Chand and the IAAF presented their case, and when we broke up, all of us were convinced that the Court would go for the IAAF line, prohibiting high testosterone woman from competing in international competitions. But three months later, CAS suspended the regulation for two years, awaiting proof from IAAF that there is a quantitative correlation between testosterone levels and physical performance. The experiment to prove such a causative effect of testosterone would be to take a group of female athletes, randomize them to active compound and placebo, and give them different amounts of testosterone and see how they perform. Such experiments would not be ethically acceptable, if enough testosterone to reach male levels would be the goal. A correlation between testosterone levels and performance in healthy elite female athletes was published in 2018, and the regulations were re-introduced, but IAAF accepted to use these only for running events 400 – 800 meters. Stephan Bermon had meanwhile proved that the correlation was valid for these events, and for 1500 meter. The latter was not included into the new regulations, since no hyperandrogenic athletes have so far been detected among the 1500 m runners. This new regulation is in effect since 2018. Presently, testosterone measurements are part of the anti-doping testing; all athletes of international competitions (World championship or Olympics) will have their testosterone levels tested. It is now easier to pick up female competitors with high testosterone levels in the male range.

Many of the athletes that after 2011 accepted to suppress their testosterone levels ended the medication after the CAS 2015 suspension of the regulations. Their athletic performance deteriorated during suppression but improved again following cessation, only to get worse again during renewed suppressive therapy.

If androgenic and anabolic hormones are administrated to athletes, it is punished as a doping violation. The effects of administered and endogenous testosterone are the same.

Strong feelings have been expressed about the IAAF regulations, both by lay people, scientists and ethicists. It is difficult to predict what direction the discussion on the regulations will take.

Ze’ev Hochberg: But Martin, is it true that females on oral contraceptives perform less well than without contraceptive?

Ritzén: Yes, we can compare the results in competition during the periods before 2011, during regulation until 2015, then a period of no regulation for 2 years after 2015, and there is a definite correlation. Semenya is the best case maybe; she was performing less well under treatment.

Saenger: But I don’t think this is a good scientific argument. To me, it is disappointing to bring lawyers that can sue for everything.

Ritzén: I am not basing this on Semenya only; all women that were found to have high levels of testosterone, before and after suppression, had better performance before suppressing testosterone. But is it not 100 cases. I think it’s pretty much proven by now.

Saenger: You talked about gender fluidity, the time is going to come when all of this gets thrown out, as there are now 3 genders in Germany, so man/woman separation does not work anymore, it becomes a challenge without the lawyers I hope, it’s clear the time will come when these things are not easily separable.

Ritzén: No, it is definitely not easily categorized. But now we’re touching on transsexuals.

Jan Lebl: Just a brief question, was a study done among male athletes? There should be a correlation between testosterone levels and performance, as there is a wide range of testosterone levels in man.

Ritzén: No, it is not that good correlation in males, and one reason may be that if a man trains very hard he will suppress his testosterone; for example, marathon runners often have low levels of testosterone.

Audience: Unless they run behind a woman!

Audience: They used to say that epitestosterone/testosterone ratio would clarify this, because in endogenous testosterone production you make both, and when you inject, there is only testosterone, and the ratio is off, so this is how you would pick out cheaters; is this correct?

Ritzén: This ratio is used in anti-doping work. However, it varies depending on ethnicity.

Arnhold: As pediatric endocrinologist, we have to take care and help patients with DSD, previously called disorders of sex development, and now it is more politically correct to say differences of sex development. Dealing with DSDs, we learned that it is a defect of nature, and we do the best that we can, but sometimes it is impossible to have an optimal result. Individuals with DSD are overrepresented in the sports field, where they are more frequent that in the general population. It is a problem for a person who was born female, lived and raised as female, excels in her sport, is a national hero, and then is told that she cannot compete anymore, or that she has to artificially lower her hormones to compete. This is a challenge.

Ritzén: I think that is the remaining difficult question: is it ethically acceptable to prevent someone from competing because she was born with a certain condition? Is it acceptable to request DSD athletes to suppress their testosterone levels, although they have not cheated, only taking advantage of their natural hormonal situation? Administration of anabolic steroids would be cheating, but she has not done so. Why should she not be allowed to compete among women? On the other hand: Is it fair to other female athletes to make them compete against women who have a male body (when it concerns muscles and haemoglobin levels)?

Arnhold: But then you also compete with other people who are taller than you, have better nutrition and socio-economic backgrounds. So there are many other factors involved.

Ritzén: But these factors are not distinguishing between male and female. Within the two categories, tallness is prevalent in both male and female. So it is not the natural definition between these two categories.

Arnhold: I looked at male and female finish times at the 800m race in the 2009 IAAF Championship: the time difference between the first male and first female (Semenya) was 10 seconds or 10%. However, the difference between Semenya and the second female was only 2 seconds or 2%. On the same race the men ran 10% faster.

Ritzén: 2 to 3 % means the difference between no 1 and no 20 in a competition. That’s the problem. Even 2% is enough in this context.

Arnhold: We know that androgen sensitivity is very difficult to measure so we don’t have an answer. I know that Caster lost an appeal. The limit of testosterone that was acceptable dropped from 10 nmol/L to 5 nmol/L.

Ritzén: The reason for the 10 nmol is in PCOS. There is a Canadian study studying the testosterone levels in woman with PCOS, with a pre-requisite of virilisation to accept the diagnosis. They had a mean levels of 3.5 nmol/L: so +5 SD in that group was 9 nmol/L. 5 SD means 1 in 10,000 or something like that. Now, the new limit, 5 nmol/L, is based on studies by Bermon, looking at 3 different world championships, reaching 3 SD above the normal female mean, so they ended up with 5 nmol/L. Also, it is based on tandem mass spectrometry, rather than immunoassay.

Ken Ong: So I understand and accept that T to determine eligibility is far better than karyotype and SRY: but do you think it is sufficient? One question: 6 months of low T may not overcome all the benefits of a lifetime high T for 20 years? Even if CAIS is really complete, such individuals have normal male birth weight and male adult height. Do you think Testosterone is sufficient?

Ritzén: It is not the only factor. It is fascinating that, in the Olympic Games, there is a clear overrepresentation of the females that have been castrated early on during childhood. It looks like the Y chromosome in itself added something beneficial to the sports. We can’t cover everything; T is a factor that can be measured.

Jan-Maarten Wit: What about 17 ketoreductase deficiency? I would assume that some of your patients/athletes that perform very well would have that diagnosis: did you see many cases?

Ritzén: No we are not seeing many cases of those; they will not be caught because they don’t have high T level. They have high androstenedione levels.

Wit: But they do perform better in sports.

Ritzén: Yes, because androstenedione is also an anabolic steroid. But we can’t cover everything.

Wit: There were countries like East Germany where doping was the rule: wouldn’t it be possible to get their data and have answers to your questions? Let me just add one more issue that shows the problem: now, we also have handicapped Olympics, and we wonder how they could be compared with such different phenotypes: is there also someone investigating like gender in that area?

Ritzén: I think they are treated the same as any other competitive athletes. The first question can be answered: there is a publication (Francke and Berendonk, Clinical Chemistry 43:7;1262–1279 (1997), who reported on studies by scientists in DDR that describe doping experiments. Women athletes were given anabolic steroids or testosterone and their sports performance was measured before and during doping. One conclusion by the doping scientists was that they did not believe that anyone could win a gold medal without being doped!

Audience: At the time it was true (laugh).

Ritzén: There are diagrams showing the sky rocking difference of times between before and after doping. Some of the protocols were burned.

Hochberg: Often times there are extreme physical differences between professional athletes and normal people: I can imagine someone with mutations that give him for instance syndactyly and he is a better swimmers, or people with genes that give them stronger muscles. Imagine people like professional American football players, they must have some special genes: would you agree to play against them?

Ritzén: There are many features that make you perform better in sports, like a tall stature, but it does not discriminate between male and female, and we are only discussing differences between males and females. Sports is divided into female sport and male sports.

Arnhold: From a medical perspective, I can see that a female patient would want to lower her T levels for her physical attributes. But for a competitive athlete, some of them risk their lives and would do anything to get better performance.

Ritzén: Many athletes benefit from having high T levels and benefit from excelling in sports. Take Caster, she comes from a poor village in South Africa and now is making a fortune, so she would not want to lower her T levels because she does not want to lower her performance. This comes back to the ethics.

Arnhold: Caster apparently didn’t know about her condition until she was excelling in sport. If she and other patients would be diagnosed early and receive the appropriate treatment, this would not happen.

Ritzén: It’s an ethical dilemma. Shall we prevent her making money and pursue her career?

Ong: Do you apply the same criteria to transsexuals?

Ritzén: The discussion for transsexuals in sports was brought up in 2004; it was discussed by a consensus committee of five people, appointed by IOC. I was one of them. The resulting recommendation was that If you transfer from male to female, you have to wait for two years after castration. At that time, castration was the rule. Now, in Sweden and many other countries, individuals can go to the authorities and say: “now I’m a woman”, without any medical workup at all. This poses a big challenge for the sports authority.

Ong: The two years has not been challenged?

Ritzén: The reason behind the 2 years is because in a study from the group of Gooren, it was shown that muscle mass decreases during the first 15 months after castration. After that period, it remains about the same.

Lyne Chiniara: I care for trans patients at Sainte-Justine hospital, and we see less and less trans patients getting gonadectomized; what about these patients that are on androgen blockers, for instance spironolactone; their T levels would still be elevated but would not have any impact because of the medication; would they be banned from competition? What happens in this situation?

Ritzén: The regulations are being reviewed in order to get a proper scientific background, but it is not yet decided on how they should be managed if they keep their testes in place.

Audience: Last question, I know it is not the topic, but what about growth hormone?

Ritzén: There are anabolic effects of GH and many athletes dope themselves with GH. What do you think? Did you find much difference in your studies?

Audience: Highly trained athletes showed no additional effects of GH.

Ritzén: There may be a synergetic effects of anabolic steroids and GH. This has been speculated but not proven. Athletes take all kind of products. GH is now in the gym, in body-building to grow muscles and lower fat mass. They take testosterone and growth hormone together.

Michael Ranke: I have looked in the composition of GH preparations and found that there is little GH in those products. This was a Russian GH and looking at the price there should be little GH in there!

Ritzén: Unfortunately, even high quality hGH has been shown to be used as doping agents. Many athletes seem to believe in its anabolic features.